Ultrasound diagnostic device, measurement condition setting method and program

Abstract

To provide an ultrasonic diagnostic apparatus, measurement condition setting method and program which can quickly set a measurement condition to ultrasonic image data.SOLUTION: An ultrasonic diagnostic apparatus comprises: an ultrasonic image generation unit 112 which generates ultrasonic image data of a subject; an inference unit 208 which infers a measurement condition candidate for the ultrasonic image data of the subject by using a learned model learned with a measurement condition set to the ultrasonic image data as teacher data; and a measurement condition setting unit 114 which sets a measurement condition to the ultrasonic image data of the subject by using the measurement condition candidate set in the inference unit 208.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to an ultrasonic diagnostic apparatus, a measurement condition setting method, and a program for performing various measurements on ultrasonic image data.

Background Art

[0002] In ultrasonic diagnostic apparatuses, it may be necessary to measure, for example, the distance between two points, the area of a region, the volume, etc. with respect to ultrasonic image data captured.

[0003] Patent Document 1 discloses searching for the start phase and end phase of a Doppler waveform and performing measurements based on the Doppler waveform. And it is disclosed that learned data stored in a learned data storage unit is used in the search.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] Patent Document 1 discloses specifying the phase of a Doppler waveform using learned data and performing measurements based on the Doppler waveform, but does not disclose setting measurement conditions using learned data and performing various measurements.

[0006] Therefore, an object of the present invention is to provide an ultrasonic diagnostic apparatus capable of quickly performing various measurements (such as distance, perimeter, area, volume, etc.) by setting measurement conditions for ultrasonic image data of a subject.

Means for Solving the Problems

[0007] In order to achieve the object of the present invention, a generation unit that generates ultrasonic image data of a subject, and using a learned model learned with measurement conditions set for the ultrasonic image data as teacher data, for the ultrasonic image data of the subject classify the measurement sites and, for the classified measurement sites, a plurality of an inference unit that infers measurement condition candidates; a measurement condition setting unit that sets measurement conditions for the ultrasonic image data of the subject by selecting one measurement condition candidate from a plurality of measurement condition candidates output from the inference unit; and a measurement unit that performs measurement on the ultrasonic image data of the subject based on the measurement conditions set by the measurement condition setting unit, wherein the measurement conditions include a measurement range at the measurement site of the subject .

Advantages of the Invention

[0008] According to the present invention, it is possible to quickly perform various measurements by setting measurement conditions for ultrasonic image data of a subject.

Brief Description of the Drawings

[0009]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Modes for Carrying Out the Invention

[0010] Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

Examples

[0011] Figure 1 shows the configuration of the ultrasonic diagnostic apparatus of the present invention. The ultrasonic diagnostic apparatus includes an ultrasonic probe 100 that contacts a subject to transmit and receive ultrasonic waves, an apparatus main body 102 that processes the ultrasonic signals received by the ultrasonic probe 100 to generate ultrasonic image data and performs various measurements, an operation unit 104 for operating the apparatus main body 102, and a display unit 106 for displaying the ultrasonic image data, measurement results, and the like.

[0012] The ultrasonic probe 100 is connected to the apparatus main body 102. The ultrasonic probe 100 has a plurality of vibrators, and can generate ultrasonic waves by driving the plurality of vibrators. The ultrasonic probe 100 receives the reflected waves from the subject and converts them into electrical signals. The converted electrical signals are transmitted to the apparatus main body 102.

[0013] Further, the ultrasonic probe 100 is provided on the front side (subject side) of the plurality of vibrators, and includes an acoustic matching layer that matches the acoustic impedances of the plurality of vibrators and the subject, and a backing material that is provided on the back side of the plurality of vibrators and prevents the propagation of ultrasonic waves from the plurality of vibrators to the back side.

[0014] The ultrasonic probe 100 is detachably connected to the apparatus main body 102. The types of the ultrasonic probe 100 include a linear type, a sector type, a convex type, a radial type, a three-dimensional scanning type, etc., and the operator can select the type of the ultrasonic probe 100 according to the imaging application.

[0015] The apparatus main body 102 includes a transmission / reception unit 110 that transmits and receives ultrasonic waves to and from the ultrasonic probe 100, an ultrasonic image generation unit that generates ultrasonic image data using the ultrasonic signals received by the transmission / reception unit 110, a measurement condition setting unit 114 that sets measurement conditions for performing various measurements, a measurement unit 116 that performs various measurements on the ultrasonic image data using the measurement conditions set by the measurement condition setting unit 114, and a control unit 118 that controls various components of the apparatus main body 102.

[0016] The transmitting and receiving unit 110 controls the transmission and reception of ultrasonic waves performed by the ultrasonic probe 100. The transmitting and receiving unit 110 includes a pulse generation unit, a transmission delay circuit, etc., and supplies a drive signal to the ultrasonic probe 100. The pulse generation unit repeatedly generates rate pulses at a predetermined repetition frequency (PRF). Also, the transmission delay circuit focuses the ultrasonic waves generated from the ultrasonic probe 100 and gives a delay time for determining the transmission directivity to the rate pulses generated by the pulse generation unit. The transmission delay circuit can control the transmission direction of the ultrasonic waves transmitted from the vibrator by changing the delay time given to the rate pulses.

[0017] In addition, the transmitting and receiving unit 110 includes an amplifier, an A / D conversion unit, a reception delay circuit, an addition unit, etc. Various processes are performed on the reflected wave signal received by the ultrasonic probe 100 to generate an ultrasonic signal. The amplifier amplifies the reflected wave signal for each channel and performs gain correction processing. The A / D conversion unit performs A / D conversion on the gain-corrected reflected wave signal. The reception delay circuit gives a delay time to determine the reception directivity to the digital data. The addition unit performs an addition process on the reflected wave signals to which the delay time has been given by the reception delay circuit. By the addition process of the addition unit, the reflection component from the direction corresponding to the reception directivity of the reflected wave signal is emphasized.

[0018] When the transmitting and receiving unit 110 two-dimensionally scans the subject, it causes the ultrasonic probe 100 to transmit two-dimensional ultrasonic waves. Then, the transmitting and receiving unit 110 generates a two-dimensional ultrasonic signal from the two-dimensional reflected wave signal received by the ultrasonic probe 100. Also, when the transmitting and receiving unit 110 three-dimensionally scans the subject, it causes the ultrasonic probe 100 to transmit three-dimensional ultrasonic waves. Then, the transmitting and receiving unit 110 generates a three-dimensional ultrasonic signal from the three-dimensional reflected wave signal received by the ultrasonic probe 100.

[0019] The ultrasonic image generation unit 112 performs various signal processes on the ultrasonic signals output from the transmission / reception unit 110 to generate ultrasonic image data. The ultrasonic image generation unit 112 performs signal processes such as detection processing and logarithmic compression on the ultrasonic signals to generate ultrasonic image data (B-mode image data) in which the signal intensity is expressed by the brightness of the luminance.

[0020] The ultrasonic image generation unit 112 can generate blood flow image data by the color Doppler method called the color flow mapping method (CFM). In the color Doppler method, ultrasonic waves are transmitted multiple times in the same direction, and frequency analysis based on the Doppler effect is performed on the received reflected wave signals to extract the motion information of the blood flow. The ultrasonic image generation unit 112 uses the color Doppler method to generate blood flow information such as average velocity, variance, and power as blood flow image data. Note that the ultrasonic image generation unit 112 may generate blood flow image data by the power Doppler method.

[0021] The measurement condition setting unit 114 sets measurement conditions for performing various measurements. The measurement conditions are at least one of the measurement site of the subject, the measurement item at the measurement site of the subject, the measurement range at the measurement site of the subject, and the like. Measurement conditions for performing various measurements are set according to the characteristics of the ultrasonic image data.

[0022] The measurement unit 116 performs various measurements on the ultrasonic image data using the measurement conditions set by the measurement condition setting unit 114. The various measurements include the distance between two points at the measurement site, the perimeter length of the measurement site, the area at the measurement site, the volume, and the like.

[0023] The operation unit 104 includes a mouse, a keyboard, buttons, a panel switch, a touch command screen, a foot switch, a trackball, a joystick, and the like. The operation unit 104 receives various instructions from the operator of the ultrasonic diagnostic apparatus and transmits the received various instructions to the apparatus main body 102.

[0024] The display unit 106 displays a GUI for the operator of the ultrasonic diagnostic apparatus to input various instructions using the operation unit 104, and also displays ultrasonic image data, blood flow image data, measurement results, etc. generated in the apparatus main body 102.

[0025] When the measurement conditions are set by the measurement condition setting unit 114, the display unit 106 displays the measurement conditions and the measurement results for the ultrasonic image data generated by the ultrasonic image generation unit.

[0026] Note that the transmission / reception unit 110, the ultrasonic image generation unit 112, the measurement condition setting unit 114, the measurement unit 116, and the control unit 118 in the apparatus main body 102 may be configured by hardware such as an integrated circuit, or may be a program modularized by software.

[0027] In a general ultrasonic diagnostic apparatus, the operator manually sets the measurement conditions via the operation unit 104 while checking a specific region in the ultrasonic image data. The ultrasonic diagnostic apparatus of the present invention can set the measurement conditions for newly generated ultrasonic image data using a learned model learned with the measurement conditions set for the ultrasonic image data as teacher data. The learned model may be, for example, a learned neural network, but any model such as deep learning or a support vector machine may be used. The learned model may be stored in the measurement condition setting unit 114, or may be stored in an information processing apparatus connected to the ultrasonic diagnostic apparatus via a network.

[0028] Specifically, the control unit 118 learns the measurement conditions set by the measurement condition setting unit 114 for the ultrasonic image data as teacher data, and generates a learned model. The control unit 118 uses the learned model to identify the cross-section of the newly generated ultrasonic image data and infer measurement condition candidates. The control unit 118 transmits the measurement condition candidates inferred by it to the measurement condition setting unit 114, and the measurement condition setting unit 114 sets the measurement conditions. The measurement conditions set in the measurement condition setting unit 114 are set for the newly generated ultrasonic image data. The measurement unit 116 performs various measurements on the ultrasonic image data using the measurement conditions set in the measurement condition setting unit 114.

[0029] Note that the ultrasonic image data input to the measurement condition setting unit 114 may be two-dimensional image data or three-dimensional image data (volume data).

[0030] As described above, the learned model is learned to infer measurement conditions. A model that identifies a cross-section from two-dimensional image data and sets measurement conditions may be used, or a model that identifies a plurality of cross-sections from three-dimensional image data and sets measurement conditions may be used.

[0031] FIG. 2 is a diagram showing the configuration of the measurement condition setting unit 114. The measurement condition setting unit 114 includes a measurement site setting unit 120, a measurement item setting unit 122, and a measurement range setting unit 124.

[0032] The measurement site setting unit 120 sets a measurement site for the ultrasonic image data generated by the ultrasonic image generation unit. Examples of the measurement site include the abdomen, chest, heart, carotid artery, fetus, etc. The measurement site setting unit 120 can also identify and set the measurement site from the ultrasonic image data generated by the ultrasonic image generation unit.

[0033] The measurement item setting unit 122 sets measurement items corresponding to the measurement site. For example, if the measurement site is the carotid artery, items for measuring the blood vessel diameter and IMT are set. The measurement item setting unit 122 can also set a gate for performing blood flow measurement (Doppler measurement).

[0034] The measurement range setting unit 124 sets a measurement range (cursor: straight line, curve, etc.) corresponding to the measurement item set by the measurement item setting unit 122 in the ultrasonic image data.

[0035] The measurement range is a measurement caliper for measuring the dimensions of the tissue depicted in the ultrasonic image data. The measurement caliper is used to specify the measurement range when measuring the measurement target. For example, by arranging the measurement caliper so as to sandwich the blood vessel displayed in the ultrasonic image data, the diameter of the blood vessel can be measured.

[0036] Generally, the measurement caliper is made to move on the screen according to the movement of the trackball of the operation unit 104 or the like. The operator aligns the measurement caliper with the measurement range of the measurement site. Then, by performing a confirmation operation, the operator can obtain the size (for example, distance, perimeter, or area) of the measurement range.

[0037] Next, the details of the control unit 118 of the present invention will be described. The configurations of the control unit 118 in FIGS. 3 and 4 are the same. The drawings are made different to distinguish the operations in the learning phase and the inference phase. FIG. 3 shows the operation of the control unit 118 related to the learning phase, and FIG. 4 shows the operation of the control unit 118 related to the inference phase.

[0038] The control unit 118 includes a learning device 200 that learns the measurement conditions when various measurements are performed on the ultrasonic image data as teacher data and generates a learned model, a storage unit 206 that stores the learned model generated in the learning device 200, and an inference unit 208 that uses the learned model to identify the measurement site in the ultrasonic image data and infer candidates for the measurement conditions for the measurement site.

[0039] The learning device 200 includes a teacher data generation unit 202 that generates teacher data regarding measurement conditions set for ultrasonic image data, and a learning unit 204 that learns the measurement conditions in the ultrasonic image data using the teacher data generated by the teacher data generation unit 202.

[0040] The ultrasonic image generated by the ultrasonic image generation unit 112 and the measurement conditions set by the measurement condition setting unit 114 are input to the learning device 200 (teacher data generation unit 202). Here, a plurality of ultrasonic image data captured in the past and the respective measurement conditions set for the plurality of ultrasonic image data are stored in the memory of the learning device 200. The teacher data generation unit 202 generates teacher data by setting (associating) the ultrasonic image data and the measurement conditions as a set. The teacher data generation unit 202 stores the set (associated) ultrasonic image data and measurement conditions in the memory. The learning device 200 (learning unit 204) learns the respective measurement conditions in the plurality of ultrasonic image data stored in the memory as teacher data.

[0041] The learning device 200 (learning unit 204) can also learn teacher data for classifying the measurement sites of the ultrasonic image data. The learning device 200 (learning unit 204) performs learning processing based on teacher data in which a correct label and a correct image are paired. As the correct label, information indicating the measurement site of the subject is set. For example, a correct label of "carotid artery" is assigned to the ultrasonic image data (correct image) of the carotid artery. A correct label of "abdomen" is assigned to the ultrasonic image data (correct image) of the abdomen. A correct label of "fetal head" is assigned to the ultrasonic image data (correct image) of the fetal head.

[0042] In this way, the learning device 200 (learning unit 204) generates a learned model (first learned model) by learning while associating the measurement site of the subject with respect to the ultrasonic image data as teacher data.

[0043] The freeze button on the operation unit 104 is a button for freezing (stopping) the ultrasonic image data being displayed in real time. When the operator presses the freeze button without moving the ultrasonic probe 100, the ultrasonic image data being displayed in real time on the display unit 106 can be frozen. The frozen ultrasonic image data can be stored in the ultrasonic diagnostic apparatus.

[0044] Here, when the operator presses the freeze button on the operation unit 104 and measurement conditions are set for the frozen ultrasonic image data on the display unit 106, the frozen ultrasonic image data and the measurement conditions displayed on the display unit 106 are output to the learning device 200. This is because the frozen ultrasonic image data is still image data suitable for the learning of the learning device 200.

[0045] Also, the learning device 200 can determine whether measurement conditions are set for the ultrasonic image data frozen on the display unit 106 based on the presence or absence of a measurement range (cursor) displayed on the ultrasonic image data.

[0046] For example, when a measurement range (cursor) is set so as to sandwich a blood vessel displayed in the ultrasonic image data, the learning device 200 can determine that measurement conditions are set for the ultrasonic image data. Then, the learning device 200 can associate and store the ultrasonic image data with the measurement conditions (measurement range), and learn the measurement conditions in the ultrasonic image data as teacher data.

[0047] The learning unit 204 uses, for example, a neural network and includes a plurality of layers. The plurality of layers have a plurality of intermediate layers between the input layer and the output layer. Although not shown in the figure, the plurality of intermediate layers include a convolutional layer, a pooling layer, an upsampling layer, and a synthesis layer. The convolutional layer is a layer that performs a convolution process on a group of input values. In the convolutional layer, convolution is performed on the input ultrasonic image data and the measurement conditions (measurement range), and the features of the ultrasonic image data and the measurement conditions (measurement range) are extracted.

[0048] The pooling layer is a layer that performs a process of reducing the number of output value groups to be less than the number of input value groups by thinning out or synthesizing the input value groups. The upsampling layer is a layer that performs a process of increasing the number of output value groups to be more than the number of input value groups by replicating the input value groups or adding interpolated values to the input value groups. The synthesis layer is a layer that inputs value groups such as the output value group of a certain layer or the pixel value group constituting the ultrasonic image data and the measurement conditions (measurement range) from a plurality of sources, and performs a process of synthesizing them by concatenating or adding them. The number of intermediate layers can be changed at any time according to the learning content.

[0049] In this way, the learning device 200 (learning unit 204) uses a neural network to generate a learned model (second learned model) by associating the measurement conditions for the ultrasonic image data as teacher data and performing learning. The first learned model and the second learned model may be generated as different models.

[0050] FIG. 5 shows an example of ultrasonic image data output to the learning device 200. FIG. 5(a) shows ultrasonic image data in which measurement ranges 402 and 404 are set for the longitudinally cut carotid artery 400. Note that the carotid artery 400 in the ultrasonic image data may have blood flow image data generated by the color Doppler method. Here, a case is shown where IMT measurement is performed in the measurement range (region of interest) 402 and distance measurement is performed in the measurement range (straight line) 404.

[0051] The IMT measurement is a method for measuring the intima-media complex in the carotid artery 400. As shown in Fig. 5(b), in the IMT measurement, the intima-media complex thickness, which is the thickness of the complex formed by combining the intima and media that make up the blood vessel wall of the carotid artery 400, is measured. The measurement unit 116 measures the IMT measurement value in the region of interest 402. The measurement unit 116 measures the average value of the distance between the inner boundary of the intima and the boundary between the middle and outer membranes in the region of interest 402 as the IMT measurement value.

[0052] The distance measurement is a method for measuring the distance between two points corresponding to the blood vessel diameter of the carotid artery 400. The measurement unit 116 measures the distance between the upper blood vessel wall and the lower blood vessel wall of the carotid artery on the straight line 404.

[0053] The learning unit 204 learns the measurement conditions (measurement site: carotid artery, measurement item: blood vessel diameter, measurement range: straight line) for the ultrasonic image data (carotid artery 400) as teacher data. In addition, the learning unit 204 learns the measurement conditions (measurement site: carotid artery, measurement item: IMT measurement, measurement range: region of interest) for the ultrasonic image data (carotid artery 400) as teacher data.

[0054] The measurement range 402 used as teacher data may be information indicating the coordinates of the region in the ultrasonic image data (for example, the coordinates of four points). In addition, the measurement range 402 used as teacher data may be the coordinates input to the operation unit 104 for setting the measurement range 402.

[0055] The measurement range 404 used as teacher data may be information indicating the coordinates of the straight line in the ultrasonic image data (for example, the coordinates of two points). In addition, the measurement range 404 used as teacher data may be the coordinates input to the operation unit 104 for setting the measurement range 404.

[0056] In this way, the learning unit 204 can learn the characteristics of the measurement conditions actually set for the ultrasonic image data. The characteristics of the measurement conditions include the measurement site, the measurement item (type of measurement), and the measurement range.

[0057] Note that the learning device 200 may be installed outside the ultrasonic diagnostic device. FIG. 6 shows an example in which the learning device 200 is installed outside the ultrasonic diagnostic device.

[0058] The learning device 200 may be, for example, in a hospital network or in a cloud outside the hospital. The learning device 200 is connected to a plurality of ultrasonic diagnostic devices 500, 502, and 504. Here, a form in which there are three ultrasonic diagnostic devices is shown, but there may be four or more ultrasonic diagnostic devices.

[0059] For example, the learning device 200 learns measurement conditions for ultrasonic image data captured by the ultrasonic diagnostic device 500 as teacher data and generates a learned model. Further, the learning device 200 learns measurement conditions for ultrasonic image data captured by an ultrasonic diagnostic device 502 different from the ultrasonic diagnostic device 500 as teacher data and updates the learned model. Similarly, the learning device 200 learns measurement conditions for ultrasonic image data captured by an ultrasonic diagnostic device 504 different from the ultrasonic diagnostic devices 500 and 502 as teacher data and updates the learned model. The learned model generated (updated) by the learning device 200 is transmitted to each of the plurality of ultrasonic diagnostic devices 500, 502, and 504. The plurality of ultrasonic diagnostic devices 500, 502, and 504 each store the latest learned model generated by the learning device 200.

[0060] In this way, the learning device 200 can learn the measurement conditions set in the plurality of ultrasonic diagnostic devices 500, 502, and 504 as teacher data. Therefore, the learning device 200 can generate a learned model corresponding to the plurality of ultrasonic diagnostic devices 500, 502, and 504.

[0061] Using FIG. 4, the measurement unit 116 regarding the inference phase will be described. The storage unit 206 is connected to the learning device 200. The storage unit 206 stores a learned model that has been trained to set measurement conditions for ultrasonic image data. Specifically, the storage unit 206 stores a learned model that has been trained to identify a specific region (such as the blood vessels of the carotid artery) from ultrasonic image data and set measurement conditions for the specific region.

[0062] The ultrasonic image data newly generated by the ultrasonic image generation unit is output to the inference unit 208. The inference unit 208 uses a learned model that has been trained to set measurement conditions for ultrasonic image data to infer measurement condition candidates for the newly generated ultrasonic image data.

[0063] The inference unit 208 uses a learned model (first learned model) that has been trained to set the measurement site of ultrasonic image data to infer the measurement site for the newly generated ultrasonic image data. Specifically, the storage unit 206 stores a learned model based on teacher data (ultrasonic image data) classified into a plurality of measurement sites (abdomen, chest, heart, carotid artery, fetus, etc.). Therefore, when new ultrasonic image data is input, the inference unit 208 can classify the measurement site for the new ultrasonic image data based on the feature amount of the new ultrasonic image data and the teacher data.

[0064] Furthermore, the storage unit 206 stores a learned model (second learned model) based on teacher data regarding measurement items and measurement ranges corresponding to a plurality of measurement sites (abdomen, chest, heart, carotid artery, fetus, etc.). Therefore, when new ultrasonic image data is input, the inference unit 208 can infer measurement condition candidates for the measurement items and measurement ranges corresponding to the measurement site of the new ultrasonic image data.

[0065] In this way, the inference unit 208 uses the learned models (the first learned model and the second learned model) to identify the measurement site for the newly generated ultrasonic image data and infer the measurement range corresponding to the measurement site.

[0066] Next, the display form of the display unit 106 of the ultrasonic diagnostic apparatus will be described with reference to FIG. 7. An inference setting button 604 for inferentially setting the measurement conditions, a manual setting button 606 for manually setting the measurement conditions, and a determination button 608 are displayed. The inference setting button 604, the manual setting button 606, and the determination button 608 correspond to the operation unit 104. The inference setting button 604, the manual setting button 606, and the determination button 608 are displayed on the display unit 106 as icons, and the operator can select either the inference setting button 604 or the manual setting button 606. Note that, as an initial setting, the inference setting button 604 may be pressed.

[0067] When the operator presses the inference setting button 604, the inference unit 208 infers measurement condition candidates for the ultrasonic image data 600 displayed on the display unit 106 using a learned model trained to set measurement conditions for the ultrasonic image data. The measurement condition candidates inferred by the inference unit 208 are output to the display unit 106 and the measurement condition setting unit 114.

[0068] The display unit 106 displays measurement condition candidates 614, 624, 634 (measurement range: straight line) for measuring the blood vessel diameter with respect to the ultrasonic image data (carotid artery 400). Here, it is assumed that three measurement condition candidates 614, 624, 634 for measuring the blood vessel diameter inferred by the inference unit 208 are displayed. The operator selects the measurement condition candidate suitable for measuring the blood vessel diameter using the indication mark 650. Then, the operator presses the determination button 608 to determine the selection. That is, the ultrasonic diagnostic apparatus of the present invention includes a selection means for selecting one measurement condition candidate from a plurality of measurement condition candidates inferred by the inference unit 208.

[0069] In FIG. 7, it shows that the measurement condition candidate 614 is selected from among a plurality of measurement condition candidates 614, 624, and 634 using the indication mark 650. The operator can select the measurement condition candidate 614 suitable for measuring the blood vessel diameter from among the plurality of measurement condition candidates 614, 624, and 634. The measurement condition setting unit 114 can set the measurement condition (measurement range: straight line) based on the selected measurement condition candidate 614.

[0070] In addition, when there is one measurement condition candidate inferred by the inference unit 208, the measurement condition setting unit 114 sets the measurement condition (measurement range: straight line) corresponding to the one measurement condition candidate. When the measurement condition candidate inferred by the inference unit 208 is not suitable, the operator presses the manual setting button 606 to manually set the measurement condition for the measurement condition setting unit 114. In this way, the form in which the measurement condition setting unit 114 sets the measurement condition differs according to the number of measurement condition candidates inferred by the inference unit 208.

[0071] The measurement unit 116 performs distance measurement using the measurement condition (measurement range: straight line 614) set by the measurement condition setting unit 114. The measurement unit 116 measures the distance between the upper blood vessel wall and the lower blood vessel wall in the carotid artery 400 under the measurement condition set by the measurement condition setting unit 114. The display unit 106 displays the measurement result of the distance measurement based on the measurement condition (measurement range: straight line 614) set by the measurement condition setting unit 114.

[0072] Also, the display unit 106 displays the measurement condition candidates 612, 622, and 632 (measurement range: region of interest) for IMT measurement with respect to the ultrasonic image data (carotid artery 400). Here, it is assumed that three measurement condition candidates 612, 622, and 632 for IMT measurement inferred by the inference unit 208 are displayed. The operator selects the measurement condition candidate suitable for IMT measurement using the indication mark 652. Then, the operator presses the determination button 608 to determine the selection.

[0073] In FIG. 7, it shows that the measurement condition candidate 632 is selected from among a plurality of measurement condition candidates 612, 622, and 632 using the indication mark 652. The operator can select the measurement condition candidate 632 suitable for IMT measurement from among the plurality of measurement condition candidates 612, 622, and 632. The measurement condition setting unit 114 can set the measurement condition (measurement range: region of interest 632) based on the selected measurement condition candidate 614.

[0074] In addition, when there is one measurement condition candidate inferred by the inference unit 208, the measurement condition setting unit 114 sets the measurement condition (measurement range: region of interest) corresponding to the one measurement condition candidate. When the measurement condition candidate inferred by the inference unit 208 is not suitable, the operator presses the manual setting button 606 to manually set the measurement condition (measurement range: region of interest) for the measurement condition setting unit 114.

[0075] The measurement unit 116 performs IMT measurement using the measurement condition (measurement range: region of interest 632) set by the measurement condition setting unit 114. The measurement unit 116 performs IMT measurement of the carotid artery 400 under the measurement condition set by the measurement condition setting unit 114. The display unit 106 displays the measurement result of the IMT measurement based on the measurement condition (measurement range: region of interest 632) set by the measurement condition setting unit 114.

[0076] The operation of the learning phase in the ultrasonic diagnostic apparatus will be described with reference to FIG. 8.

[0077] S700: The operator brings the ultrasonic probe 100 into contact with the subject. The ultrasonic probe 100 may be brought into contact with the subject via ultrasonic jelly. With the ultrasonic probe 100 in contact with the subject, the transmission / reception unit 110 transmits and receives ultrasonic waves to and from the ultrasonic probe 100.

[0078] S702: The ultrasonic image generation unit performs various signal processes on the ultrasonic signal generated by the transmission / reception unit 110 from the reflected wave signal to generate ultrasonic image data.

[0079] S704: The operator determines whether to set measurement conditions for the ultrasonic image data via the operation unit 104 (measurement condition setting unit 114). At this time, whether to set measurement conditions for the ultrasonic image data may be determined according to the measurement site. For example, when the measurement site is a predetermined measurement site such as the carotid artery or the fetus, it can also be regarded as setting measurement conditions for the ultrasonic image data via the operation unit 104 (measurement condition setting unit 114). If measurement conditions are set for the ultrasonic image data, proceed to S706; if measurement conditions are not set for the ultrasonic image data, the operation in the learning phase ends.

[0080] S706: The measurement condition setting unit 114 sets measurement conditions for the ultrasonic image data generated by the ultrasonic image generation unit. The measurement conditions set by the measurement condition setting unit 114 are transmitted to the learning device 200.

[0081] S708: The learning device 200 learns the measurement conditions set in the ultrasonic image data as teacher data and generates a learned model. The learning device 200 may also learn the ultrasonic image data and the measurement conditions as teacher data and generate a learned model. After S708, the operation in the learning phase ends.

[0082] Next, with reference to FIG. 9, the operation in the inference phase of the ultrasonic diagnostic apparatus will be described.

[0083] S800: The operator brings the ultrasonic probe 100 into contact with the subject. In a state where the ultrasonic probe 100 is in contact with the subject, the transmission / reception unit 110 transmits and receives ultrasonic waves to and from the ultrasonic probe 100.

[0084] S802: The ultrasonic image generation unit performs various signal processes on the ultrasonic signal generated by the transmission / reception unit 110 from the reflected wave signal and generates ultrasonic image data.

[0085] S804: The operator determines whether to set measurement conditions for the ultrasonic image data via the operation unit 104 (measurement condition setting unit 114). At this time, depending on the imaging site, it may be determined whether to set measurement conditions for the ultrasonic image data. When the measurement site is a predetermined measurement site such as the carotid artery or fetus, it can also be regarded as setting measurement conditions for the ultrasonic image data. When setting measurement conditions for the ultrasonic image data, proceed to S808; when not setting measurement conditions for the ultrasonic image data, proceed to S806.

[0086] S806: The display unit 106 displays only the ultrasonic image data generated by the ultrasonic image generation unit. After S806, the operation in the inference phase ends.

[0087] S808: The operator determines whether to infer and set measurement conditions via the operation unit 104. For example, as shown in FIG. 7, the operator selects either the inference setting button 604 or the manual setting button 606. When inferring and setting measurement conditions, proceed to S812; when not inferring and setting measurement conditions, proceed to S810.

[0088] S810: When the operator presses the manual setting button 606, the operator manually sets the measurement conditions via the operation unit 104 (measurement unit 116) while checking the ultrasonic image data 600.

[0089] S812: The inference unit 208 infers measurement condition candidates for the ultrasonic image data displayed on the display unit 106 using a learned model trained to set measurement conditions for the ultrasonic image data. The measurement condition candidates inferred by the inference unit 208 are transmitted to the measurement condition setting unit 114.

[0090] S814: The measurement condition setting unit 114 sets the measurement conditions from the measurement condition candidates. The display unit 106 displays the measurement results measured according to the measurement conditions generated by the measurement condition setting unit 114 for the ultrasonic image data generated by the ultrasonic image generation unit. After S814, the operation in the inference phase ends.

[0091] In this way, when the operator brings the ultrasonic probe 100 into contact with the subject and displays the ultrasonic image data, the measurement conditions are automatically set. Therefore, measurements based on the measurement conditions can also be automatically performed simultaneously with the display of the ultrasonic image data. The operator can obtain the measurement results regarding the ultrasonic image data by simply bringing the ultrasonic probe 100 into contact with the subject and pressing the freeze button.

[0092] As described above, the ultrasonic diagnostic apparatus in this embodiment includes a generation unit (ultrasonic image generation unit 112) that generates ultrasonic image data of a subject, an inference unit 208 that infers candidate measurement conditions for the ultrasonic image data of the subject using a learned model learned with the measurement conditions set for ultrasonic image data other than the subject as teacher data, and a measurement condition setting unit 114 that sets measurement conditions for the ultrasonic image data of the subject using the candidate measurement conditions set in the inference unit 208. It further includes a measurement unit 116 that performs measurements on the ultrasonic image data of the subject based on the measurement conditions set by the measurement condition setting unit 114.

[0093] Therefore, according to the present invention, by using a learned model learned with the measurement conditions set for ultrasonic image data (first ultrasonic image data) as teacher data, the measurement conditions can be quickly set for newly generated ultrasonic image data (second ultrasonic image data).

Embodiment

[0094] The ultrasonic diagnostic apparatus in Embodiment 2 of the present invention will be described with reference to FIGS. 2 to 4 and FIG. 10. The difference from Embodiment 1 is that measurement conditions are set for the ultrasonic image data obtained by photographing a fetus.

[0095] As shown in FIG. 2, the measurement site setting unit 120 sets the measurement site for the ultrasonic image data generated by the ultrasonic image generation unit. The measurement site in this embodiment is a fetus.

[0096] If the measurement site is the fetus, the measurement item setting unit 122 sets the biparietal diameter (the diameter between the left and right parietal bones), the trunk circumference (the length around the fetus's abdomen), the femur length (the length of the thigh bone), etc. as measurement items.

[0097] The measurement range setting unit 124 sets a measurement range (cursor) corresponding to the measurement item set by the measurement item setting unit 122 in the ultrasonic image data. The measurement range (cursor) is a measurement caliper for measuring the dimensions of the tissue depicted in the ultrasonic image data. The measurement caliper is used to specify the measurement range when measuring the measurement target. For example, the biparietal diameter can be measured by arranging the measurement caliper so as to sandwich the head from the left and right. Also, the trunk circumference can be measured by arranging the measurement caliper so as to surround the abdomen.

[0098] The measurement unit 116 performs various measurements on the ultrasonic image data using the measurement conditions set in the measurement condition setting unit 114 (measurement site setting unit 120, measurement item setting unit 122, measurement range setting unit 124).

[0099] As shown in Figure 3, the learning unit 204 learns a plurality of measurement conditions for the ultrasonic image data (fetus) as teacher data. Here, the learning unit 204 learns the following plurality of measurement conditions as teacher data: (measurement site: fetal head, measurement item: biparietal diameter, measurement range: straight line), (measurement site: fetal abdomen, measurement item: trunk circumference, measurement range: circumference), (measurement site: fetal femur, measurement item: femur length, measurement range: straight line).

[0100] For example, when a measurement range (cursor) is set for the fetus displayed in the ultrasonic image data, the learning device 200 can determine that measurement conditions are set in the ultrasonic image data. Then, the learning device 200 can associate and store the ultrasonic image data with the measurement conditions (measurement range), and learn the measurement conditions in the ultrasonic image data as teacher data.

[0101] In this way, the learning unit 204 can learn the characteristics of the measurement conditions actually set for the ultrasonic image data. The characteristics of the measurement conditions include the measurement site, the measurement item (type of measurement), and the measurement range.

[0102] As shown in FIG. 4, the ultrasonic image data newly generated by the ultrasonic image generation unit 112 is output to the inference unit 208. The inference unit 208 uses a learned model trained to set measurement conditions for the ultrasonic image data to infer measurement condition candidates for the newly generated ultrasonic image data.

[0103] Specifically, the storage unit 206 stores a learned model (first learned model) based on teacher data (ultrasonic image data) classified into a plurality of measurement sites (fetal head, fetal abdomen, fetal thigh bone, etc.). Therefore, when new ultrasonic image data is input, the inference unit 208 can classify the measurement site for the new ultrasonic image data based on the feature amount of the new ultrasonic image data and the teacher data.

[0104] Furthermore, the storage unit 206 stores a learned model (second learned model) based on teacher data regarding measurement items and measurement ranges corresponding to a plurality of measurement sites (fetal head, fetal abdomen, fetal thigh bone, etc.). Therefore, when new ultrasonic image data is input, the inference unit 208 can infer measurement condition candidates for the measurement items and measurement ranges corresponding to the measurement site of the new ultrasonic image data.

[0105] As shown in FIG. 10, the measurement unit 116 performs measurement using the measurement conditions set in the measurement condition setting unit 114. FIG. 10(a) is an example showing the measurement site: fetal head, measurement item: biparietal diameter, and measurement range: straight line. The measurement unit 116 measures the left - right distance in the fetal head under the measurement conditions set in the measurement condition setting unit 114. The display unit 106 displays the measurement result of the distance measurement based on the measurement conditions (measurement range: straight line 900) set in the measurement condition setting unit 114 as the biparietal diameter. FIG. 10(b) is an example showing the measurement site: fetal abdomen, measurement item: trunk circumference, and measurement range: circumference. The measurement unit 116 measures the circumference in the fetal abdomen under the measurement conditions set in the measurement condition setting unit 114. The display unit 106 displays the measurement result of the circumference based on the measurement conditions (measurement range: curve 902) set in the measurement condition setting unit 114 as the trunk circumference. FIG. 10(c) is an example showing the measurement site: fetal femur, measurement item: femoral length, and measurement range: straight line. The measurement unit 116 measures the distance in the fetal femur under the measurement conditions set in the measurement condition setting unit 114. The display unit 106 displays the measurement result of the distance measurement based on the measurement conditions (measurement range: straight line 904) set in the measurement condition setting unit 114 as the femoral length.

[0106] Therefore, according to the present invention, by using a learned model learned with the measurement conditions set for ultrasonic image data (such as fetal head, fetal abdomen, fetal femur, etc.) as teacher data, the measurement conditions can be quickly set for newly generated ultrasonic image data (such as fetal head, fetal abdomen, fetal femur, etc.).

[0107] A computer program that realizes the functions of Examples 1 and 2 can be supplied to a computer via a network or a storage medium (not shown), and the computer program can be executed. It is a computer program for causing a computer to execute the above - described ultrasonic image data display method. That is, the computer program is a program for realizing the functions of an ultrasonic diagnostic apparatus on a computer. The storage medium stores the computer program.

Explanation of Reference Numerals

[0108] 100 Ultrasonic probe 102 Device body 104 Operation unit 106 Display unit 110 Transmission / reception unit 112 Ultrasonic image generation unit 114 Measurement condition setting unit 116 Measurement unit 118 Control unit 120 Measurement site setting unit 122 Measurement item setting unit 124 Measurement range setting unit 200 Learning device 202 Teacher data generation unit 204 Learning unit 206 Memory unit 208 Inference unit

Claims

1. A generator for generating ultrasound image data of a subject; an inference unit that classifies measurement sites for the ultrasound image data of the subject using a trained model trained using measurement conditions set for the ultrasound image data as teacher data, and infers a plurality of measurement condition candidates corresponding to the classified measurement sites; a measurement condition setting unit that sets a measurement condition for the ultrasound image data of the subject by selecting one measurement condition candidate from the plurality of measurement condition candidates output from the inference unit; a measurement unit that performs a measurement on the ultrasound image data of the subject based on the measurement conditions set by the measurement condition setting unit, The ultrasonic diagnostic apparatus according to claim 1, wherein the measurement conditions include a measurement range in a measurement region of the subject.

2. The ultrasound diagnostic apparatus according to claim 1 , further comprising a learning device that learns measurement conditions in ultrasound image data as teacher data and generates the trained model.

3. The ultrasound diagnostic device according to claim 2, characterized in that when measurement conditions are set for ultrasound image data frozen on the display unit, the ultrasound image data and the measurement conditions displayed on the display unit are output to a learning device.

4. The ultrasound diagnostic apparatus according to claim 2, characterized in that the learning device determines whether or not measurement conditions are set for the ultrasound image data frozen on the display unit based on the presence or absence of a measurement range displayed on the ultrasound image data.

5. The learning device uses a neural network to perform a priori processing on the ultrasound image data. The ultrasonic diagnostic apparatus according to claim 2, wherein the measurement conditions are used as training data and are learned in association with each other to generate a trained model.

6. The ultrasound diagnostic apparatus according to claim 2 , further comprising a storage unit that stores a trained model generated by the learning device.

7. The ultrasound diagnostic apparatus according to claim 1, characterized in that a learning device that learns measurement conditions in ultrasound image data as teacher data and generates the trained model is installed outside the ultrasound diagnostic apparatus.

8. The ultrasonic diagnostic apparatus according to claim 7 , wherein the learning device learns measurement conditions set in a plurality of ultrasonic diagnostic apparatuses as teacher data.

9. The ultrasound diagnostic device according to claim 1, further comprising a memory unit that stores a first trained model based on teacher data of ultrasound image data classified into a plurality of measurement regions, and a second trained model based on teacher data regarding measurement items and measurement ranges corresponding to the plurality of measurement regions.

10. 10. The ultrasound diagnostic apparatus according to claim 9, wherein the inference unit uses the first trained model to identify a measurement portion for newly generated ultrasound image data, and uses the second trained model to infer the plurality of measurement condition candidates corresponding to the measurement portion.

11. 2. The ultrasonic diagnostic apparatus according to claim 1, wherein the measurement range is a range consisting of a straight line or a curved line.

12. 2. The ultrasonic diagnostic apparatus according to claim 1, wherein the measurement site is a carotid artery, and the plurality of measurement condition candidates are measurement condition candidates for measuring a blood vessel diameter.

13. 2. The ultrasonic diagnostic apparatus according to claim 1, wherein the measurement site is a carotid artery, and the plurality of measurement condition candidates are measurement condition candidates for performing IMT measurement.

14. generating ultrasound image data of the subject; A step of classifying measurement sites for the ultrasound image data of the subject using a trained model trained using measurement conditions set for the ultrasound image data as teacher data, and inferring a plurality of measurement condition candidates corresponding to the classified measurement sites; and setting a measurement condition for the ultrasound image data of the subject by selecting one measurement condition candidate from the plurality of inferred measurement condition candidates. The measurement condition setting method, wherein the measurement conditions include a measurement range in a measurement site of the subject.

15. A program for causing a computer to execute the measurement condition setting method according to claim 14.

Images